There are three methods (sequential compensation method, collective compensation method, and best-subset selection method) to perform mixing and weighing of plural kinds, for example, first to N-th kinds of objects to be weighed, in a conventional mixing-weighing combination weigher. In these methods, predetermined are a target weight value for each kind (kind-based target weight value) and an allowable range for each kind (allowable kind-based range) for each of the first to N-th kind of objects to be weighed, and a total target weight value (target combination weight value) of all kinds and a total allowable range (allowable combination range) of all kinds. The kind-based target weight value for each kind is predetermined so that the total target weight value of all kinds becomes equal to the target combination weight value.
In the sequential compensation method, first, for the first kind, a combination weight value that is closest to a kind-based target weight value of the first kind is found. A kind-based target weight value of the second kind is compensated so as to compensate a difference between that combination weight value and the kind-based target weight value of the first kind. For the second kind, a combination weight value that is closest to the compensated kind-based target weight value of the second kind is found. This calculation is repeated to find combination weight values for respective of all kinds.
In the collective compensation method, for first kind to (N−1)-th kind, combination weight values that are closest to respective kind-based target weight values are found, and a kind-based target weight value of the N-th kind is compensated to compensate a difference between a total of the combination weight values found for the first kind to the (N−1)-th kind and a total of the kind-based target weight values of the first kind to the (N−1)-th kind. For the N-th kind, a combination weight value that is closest to the compensated kind-based target weight value of the N-th kind is found.
In the best-subset selection method, for each of the first kind to the N-th kind, combination weight values in the allowable kind-based range are all found, and are selected as combination candidates. For each of the first kind to the N-th kind, one of the combination weight values selected as the combination candidates is selected, and the selected combination weight values are combined to form combinations, and a combination in which a total of the combined combination weight values falls within an allowable combination range and is closest to the target combination weight value is found (see, for example, Patent Document 1).
Patent Document 1: Japanese Laid-Open Patent Application Publication No. Hei. 7-198466
Patent document 2: Japanese Examined Patent Application Publication No. Hei. 6-56313